interpose the same piece of iron lengthways; in which case the position of the compass needle will be scarcely or not at all affected.
The assertion of Bernoulli and others, that the absolute force of the artificial magnet increases in the ratio of its superficies, stands corrected in the far more accurate experiments of Coulomb (published in his Treatise on Magnetism), which proves that the increase takes place (in a far greater degree) in the ratio of its length. The same naturalist even found means to determine that the directing powers of the needle, which he had measured by help of his balance de tortion, stand to the length of the needle in such a ratio as that, provided only the length of the needle is from forty to fifty times its diameter, the momenta of these directing powers will increase in the very same direct proportion as the length is increased. Nor is this all that may be deduced from the experiment last mentioned. If only the magnet be strong enough, it will show likewise that magnetism seeks the length. The proof is contained in the remarkable fact, that the iron interposed between the magnet and the magnetic needle breadthways constantly acquires its two opposite poles at both ends lengthways. Though the preceding experiments are abundantly sufficient to prove the position, yet the following deserves mention for the beautiful clearness of its evidence. If the magnetic power is determined exclusively by length, it is to be expected that it will manifest no force, where the piece of iron is of such a shape that no one dimension predominates. Bring a cube of iron near the magnetic needle and it will not exert the slightest degree of power beyond what belongs to it
